ExoPlanet detection

[dave1978]

So i was listening to radio 4 last night and have seen various programmes about the technique for ExoPlanet detection using the gravity wobble observation or doppler method.

This got me thinking, surely this method which so far has been the most productive method is finding Exo's is hugely flawed.  I mean that it can only be a reliable test if the star being focused on only has 1 single Exo planet.  They cant measure the wobble of a star if there are for instance 10 Exo planets in the system.

So is this a reliable method and if so how do they get around there being more than 1 planet in the star system?

[ronin]

Sure that have found multiple planet systems using the method.

Suppose that what you get is a change in the doppler effect from the different planets and their positions. The effect would be like adding 2 different sine waves together. What is done down here is determining which 2 sine waves would give the result.

Lots of research into exoplanet detection so contact a university that performs it and ask, some may have talks on the subject that you can attend.

Think they have found one system with about 8 planets revolving around it by this method.

[Matt Scunthorpe]

I think it just identifies that there are planets in orbit rather than how many, like you say. But if there were alot of planets pulling it, then they would 'pull the star' in all directions rather, as they were all be orbiting there centre of mass. I may be wrong but this is what makes sense in my head. LOL.

Matt.

[JB80]

10 maybe not, 7 though they can.

http://www.bbc.co.uk/news/science-environment-24642603

A stars size and brightness is comparatively big when compared to the orbits of solar system objects and because they orbit around the star it's kind of like that old saying "A broken clock is right twice a day".

It just takes a bit longer than a day is all.

[dave1978]

10 maybe not, 7 though they can.

http://www.bbc.co.uk/news/science-environment-24642603

A stars size and brightness is comparatively big when compared to the orbits of solar system objects and because they orbit around the star it's kind of like that old saying "A broken clock is right twice a day".

It just takes a bit longer than a day is all.

Kepler uses the transit method for detection which i dont have issues with. 

Surely if more than 1 planet is tugging away at a star these wobbles are immeasurable as 1 conflicts the other.  So again the doppler method can only be accurate if there is a single planet.

[Knight of Clear Skies]

Surely if more than 1 planet is tugging away at a star these wobbles are immeasurable as 1 conflicts the other.  So again the doppler method can only be accurate if there is a single planet.

The method relies on taking a large number of observations to see how light from the parent star is blue and red shifted over time. Let's take the simplest example: two planets in orbit around a star. The two planets orbit at different distances, so have different orbital periods. If you were to plot the shift over time you'd get something like this:

B shows the wobble caused by a planet close to the star while A is a planet in a more distant orbit. C is the combined wobble these two planets would induce in the parent star. It's possible to split the complex graph into its component parts using a Fourier analysis.

The method is sound, but claims of seven planet systems should be treated with a touch of scepticism: what they mean is a seven planet solution is the best fit to the data they have collected. Also please bear in mind that both this technique and the transit method introduce considerable observational bias, both are far better at finding large planets in close orbits than small ones in distant orbits.

[dave1978]

I just cant get my head to accept a star can have more than 1 wobble,  you cant wobble 3 different ways at once.

[Pig]

Did they use our sun and solar system as a model ?

At least we know this to be correct.

[Knight of Clear Skies]

Mental blocks I can relate to. Things can wobble in complex motions when there are multiple forces acting on them, say from the gravitational attraction of several planets. The key is to track those movements over time, from a single reading it's not possible to detect the influence of any companions.

[LTB]

The Wobble method only works with large planets.  It is not accurate enough to detect earth like planets in orbits around distant stars primarily because earth like planets do not exhibit enough gravitational pull on their correseponding stars to affect them.  As a result the new Webb telescope is going to be launched and sent a million kilometers from earth to escape the effects of Infared from both the sun and the earth.  They are fully expecting to be able to detect earth like planets with this new telescope.

[Knight of Clear Skies]

This may help. Here's an exaggerated orbital simulation I generated from here:

As you can see, the motion of the Sun in the middle is quite complex, and couldn't be produced by the influence of a single body in orbit around it. if you want to give this a go and see it in action go to the site I linked to above, select 4 bodies and edit the masses to the same values I used (200, 20, 0.00001 and 5).

[Planetesimal]

That simulator is cool! I just programmed in a 2 body system and managed to crash the inner object into the outer one - for some reason the outer survived but the inner exploded!

[Kropster]

Sifting out the individual planets and their orbits is done using mathematical techniques which can separate the components from the observed complex wobble.

It might seem incredible, but the maths really can crunch the data and spit out a valid result.

[Zakalwe]

Kepler uses the transit method for detection which i dont have issues with. 

The transit method is only good if we are looking along the orbital plane though.  If not, then we don't see the transit.

[themos]

I just cant get my head to accept a star can have more than 1 wobble,  you cant wobble 3 different ways at once.

Actually, you can wobble an infinite number of ways at once. 

[symesie04]

I just cant get my head to accept a star can have more than 1 wobble,  you cant wobble 3 different ways at once.

Not sure how your envisioning it. It will be real hard or impossible to equate in your head if you assume circular orbits for the planets. However if the orbits are eliptical (which they most certainly will be) then its easier as their gravitational force do not pull on the star all at once (except maybe rare occasions. So if you imagine a two planet system then imagine one tugging on the star at its closest approach and then heading off onto its further reach of orbit. In the meantime the other planet heads in and tugs on the star before heading etc etc. Here there are two distinct wobbles in the stars spin and by measuring over a period of time these can be calculated into numbers, possible size, mass and orbital distance of the planets. Of course false readings can arise from unexpected patterns within a multiple planet orbit (and the more planets the trickier it gets) but generally it appears to be reasonably accurate.

[Knight of Clear Skies]

Did they use our sun and solar system as a model ?

At least we know this to be correct.

Interesting thought, sanity checking methods is always a good idea. A quick google shows that the doppler effect is seen in sunlight, but is used to measure it's rotation rate and the activity of convection cells (granules). I guess it would be a lot of work to try and extract Jupiter's orbit from the data.

That simulator is cool! I just programmed in a 2 body system and managed to crash the inner object into the outer one - for some reason the outer survived but the inner exploded!

Smashing planets together never gets old. I think the simulator deals with collisions by merging the two bodies. If you smash two planets of equal size together the resulting blob is bigger than the originals.